融合图像和指标的阿尔茨海默病多分类诊断模型

Multi classification of Alzheimer’s disease using linear fusion with images and indicators

随着人工智能的发展,计算机辅助诊断在阿尔茨海默病诊断中扮演着越来越重要的角色.本文提出了一种融合图像和指标的新型多分类诊断模型,充分挖掘TOP-MRI图像和临床指标特征用于阿尔茨海默病的多分类诊断.首先,构建由3个VGGNet-16卷积神经网络和1个单隐层网络组成的TOP-CNN-NN模型提取大脑TOP-MRI图像特征向量,利用CfsSub-setEval评估器来筛选临床指标组成指标特征向量;然后,采用典型相关分析(CCA)方法将图像特征向量和指标特征向量进行线性融合;最后,将融合特征向量输入多分类分类器来区分阿尔茨海默病的3个阶段,包括正常(CN)、轻度认知障碍(MCI)和阿尔茨海默病(AD).通过ADNI公开数据集证明,本文提出方法在阿尔茨海默病多分类诊断上的正确率可达到86.7%,有较好的性能表现.

With the development of artificial intelligence,computer-aided diagnosis plays an increasingly important role in the diagnosis of Alzheimer's disease.This paper proposes a new multi-classification diagnostic model based on images and indicators fusion.TOP-MRI image and clinical indicator features are fully exploited for multi-classification diagnosis of Alzheimer's disease.Firstly,a TOP-CNN-NN model consisting of three VGGNet-16 convolutional neural networks and a single hidden layer network is constructed to extract the feature vector of TOP-MRI images of the brain,then CfsSubsetEval evaluator is used to screen the feature vector of clinical indicators.Secondly,the image feature vector and the indicator feature vector are linearly fused by Canonical Correlation Analysis(CCA).Finally,the vector is inputted into multi-classifier,which is used to distinguish three stages of Alzheimer's disease,including normal(CN),mild cognitive impairment(MCI)and Alzheimer's disease(AD).Experiments show that the accuracy of the proposed method in multi-classification diagnosis of Alzheimer's disease can reach 86.7%,and it has good performance.